OpenAI launches Images 2.0

OpenAI rolled out ChatGPT Images 2.0 on April 21, adding a new image model with better text rendering, multilingual output, and API access. (openai.com) The release page says the model is designed to handle text inside images more accurately and to support prompts and outputs across multiple languages. OpenAI also says the model adds stronger visual reasoning, which means it can use image and text context together when generating pictures. (openai.com) On the developer side, OpenAI’s documentation says image generation is available through its API, where models can generate or edit images from prompts and image inputs. The docs now point developers to GPT Image 2 for image output workflows. (developers.openai.com 1) (developers.openai.com 2) OpenAI’s pricing page lists GPT-image-2 at $5 per 1 million text input tokens, $8 per 1 million image input tokens, and $30 per 1 million image output tokens. That token billing model differs from older flat per-image pricing that many image tools used. (openai.com) OpenAI’s older GPT Image 1 model page still shows per-image reference prices alongside token rates, including $0.167 for a high-quality 1024x1024 image and $0.25 for high-quality 1536x1024 or 1024x1536 output. That gives developers a rough benchmark for how OpenAI has been translating token usage into image costs. (developers.openai.com) The practical pitch is straightforward: a company can ask the model to produce a product card, poster, menu, or social ad in a local language without building a full design pipeline first. OpenAI’s earlier API launch for its image model said businesses could integrate professional-grade image generation directly into their own tools and platforms. (openai.com) That matters for small teams because text inside images has been one of the weak spots in image generation systems for the past two years. If the model can place legible words in the right language, it becomes more useful for commerce, marketing, and catalog work instead of just concept art. (openai.com) OpenAI has been iterating quickly on this product line. In December 2025, it introduced a new ChatGPT Images experience and said the model behind it was also available in the API as GPT Image 1.5; four months later, it is now marketing ChatGPT Images 2.0 as the next step. (openai.com 1) (openai.com 2) The company has not, in the material reviewed here, published a simple “2K mode” price card on the launch post itself. What it has published is a new product announcement, current API pricing for GPT-image-2, and developer documentation that shifts image-generation guidance toward that newer model. (openai.com 1) (openai.com 2) (developers.openai.com) For now, the release gives OpenAI a cleaner story to sell: generate images that can spell, localize, and ship through the API. Whether developers adopt it at scale will likely come down to cost, reliability, and how well those text-heavy images hold up in production. (openai.com) (openai.com)